I’ve noticed that the prints on my own 3D printer are snug compared to the same parts printed by Shapeways. Indeed, the sand dome and the ash pan wanted a little scraping where they connect with other parts. So, I made a 2 cm test cube to see what I could learn about the accuracy.
Given the discussion here, I was expecting overall shrinkage of the part. What I found is that my part was generally thicker, but not smaller. The outside dimension was 0.05 to 0.08 mm bigger than design. The inside dimension is similarly 0.04 to 0.08 mm smaller than designed.
The vertical accuracy is difficult to judge from this print because the supports were on the bottom along with some resin that failed to drain. Moreover, I made the supports too light, which allowed the edges of the print to start to separate under the weight of the cube (lesson learned!).
The other thing to note from this print is the finer surface finish at the corners than on the sides of the cube. I believe this is because the corners are well supported, compared with the sides. So, as the print built, the sides were free to wobble, while the corners were not.
if you print another cube the same way, do you get the same result? In other words, within a range, is the inaccuracy repeatable so you could redesign accordingly?
The parts tend to grow because the UV light bleeds to the next pixel slightly.
Good question, Rob.